Pinboard and Indieweb.xyz as clustering tools.
Ok ok, one other thing that has dawned on me: it’s not just the emergent connections between writers that is salient when clustering. It’s the connections between readers as well! (This is one thing that Google cannot possibly capture.)
To akaKenSmith’s point:
Having found each other, kindred parties need a work space where they can develop shared understandings.
The old Delicious was this kind of workspace for readers - a similar effort can be found in Pinboard.
One interesting thing I like to do with Pinboard is to look up a link - say ‘The Zymoglyphic Musem’ (results here) and then look at the other bookmarks for those who found the link. For example, the user PistachioRoux.
All of those links are now related to ‘The Zymoglyphic Museum’ by virtue of being in the realm of interest of PistachioRoux. YouTube uses these sorts of algorithms to find related videos by matching your realms of interest with someone else’s. However, in the process, that person is removed. (Or ‘those people’, more appropriately.) PistachioRoux is removed.
But perhaps PistachioRoux is the most interesting part of the discovery.
Particularly in a world which is becoming dominated by writers rather than readers - maybe the discovery of valuable readers is part of this.
Say a post tagged with #how_to #mk #fix_stabs could be crawled and collected into a single mechanical keyboard maintenance page. All that really calls for is emergent keywords from communities and tagging posts which bloggers can do and automations can assists with.
This does sound a lot like Indieweb.xyz, as @jgmac1106 mentioned. The concept is simple:
- Blogger ‘tags’ their post with a URL: https://indieweb.xyz/en/mk.
- Their Webmention (pingback) software notifies that URL: “Hey, a post has been made on this tag.”
- Indieweb.xyz checks the page for a valid link - sure enough.
- The blog post is added to that URL on Indieweb.xyz.
So the emergence should come from blogs clustering around a given URL.
I’ve been wondering if they could do a similar thing with http://www.adfreeblog.org/ - a ‘general’ blog community could be established around a simple ideal like that.
Might look like this:
- Blog links to adfreeblog.org on their home page.
- Adfreeblog.org notices visitors coming from that page and checks that page for the link and the image.
- If found, it adds the blog to a directory, using the meta description and keyword tags.
The adfreeblog.org home page then becomes a directory of the community. So, kind of like a webring, but actually organized. With Twitter cards and such floating in the metadata, it is probably much easier to extrapolate a good directory entry.
Spam is an issue with this approach - but it’s a start toward discovery. There aren’t a whole lot of ways for a blog to jump out from the aether and say, “I’m over here - blogging about keyboards too!” And, in a way, the efforts to squash abuse and harassment are making it more difficult.
This can become an important component in the new discovery system like how awesome-blahblah github repos are playing a key role in open source discovery.
I think it’s important to point out, though, that ‘awesome’ directories are intended to be human-curated, not generative. They feel like a modern incarnation of the old ‘expert’ pages.